Error with data envelopment analysis

[Marco Greco]

Hi everyone, I am using the dea command on Stata 12.0 (Data envelopment analysis. Stata Journal 10: 267–280) with the following specification:

  dea y1 y2 = x1 x2 x3 x4 x5, ort(out)

all input variables except x5 are binary. y1 and y2 are ratios included in 0-100 range

I am returned the error message:

 variable dmu not found
  in option rownames()
  r(111);

Can you help me to understand what is going wrong? I am using Stata 12.0

Thank you very much,
Marco

[Scott Merryman]

dea requires a variable name dmu which identifies the decision making units.

[Marco Greco]

Thank you very much for your feedback. I added the variable dmu. Now I am experiencing a different problem: from the following code:

 dea y1 y2 = x1 x2 x3 x4 x5, rts(vrs) ort(out) saving(DEAintRD)

I receive the following error:

matrix has missing values
r(504);

actually I am sure that all dmus with missing inputs or outputs have been discarded from the analysis, so I don't understand which matrix has missing values, and why!!

[Scott Merryman]

You might try Nick Cox's missings from SJ15-4 to check for missing values.

. use "cooper_table3.7.dta", clear

. replace o_profits = . in 2
(1 real change made, 1 to missing)

. dea i_employee i_area = o_sales o_profits  , rts(vrs)  ort(out)
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
      name:  dealog
       log:  F:\data\dea.log
  log type:  text
 opened on:   9 Mar 2016, 13:14:21
matrix has missing values
r(504);


. ssc inst missings, replace
checking missings consistency and verifying not already installed...
all files already exist and are up to date.

. missings report

Checking missings in all variables:
1 observation with missing values

o_profits   1

[Marco Greco]

Thanks for this, but even using this tool no missings are present among inputs and outputs of the model...